1. Forschungszentrum Informatik, Karlsruhe FZI Research Center for Information Science at the University of Karlsruhe Variance in e-Business Service Discovery Stephan Grimm, Boris Motik, Chris Preist (HP Labs, Bristol)

2. Slide 2 Overview Introduction Intuition behind modelling service semantics Operationalising discovery using logic Matching service descriptions Conclusion

3. Slide 3 Service Discovery in the Semantic Web Service – Web Service vs. high-level eBusiness Service Service Discovery – Locating Providers who meet a Requestor´s needs – Based on Semantic Descriptions of Services Semantic Description of a Service – Describing the Capabilities of the Service – Using ontology languages, such as OWL – Referring to common domain ontologies

4. Slide 4 Overview Introduction Intuition behind modelling service semantics Operationalising discovery using logic Matching service descriptions Conclusion

5. Slide 5 Service Description – Service Instance shipping 1 BremenPlymouth from to package X item 50 kg weight set of accepted Service Instances... shipping 2 HamburgDover from to barrel Y item 25 kg weight Service Instances Service Description Shipping containers from UK to Germany describes

6. Slide 6 Variance in Service Descriptions Two kinds of variance in service descriptions to shipping 2 Hamburg... to shipping 1 Bremen to shipping 3 Boston – due to incomplete knowledge... to shipping 2 Hamburg to shipping 1 Bremen Shipping to Germany – due to intended diversity different service instances... different possible worlds

7. Slide 7 Discovery by Matching Service Descriptions Matching Service Descriptions of Requestors an Providers If there are common instances, requestor and provider can (potentially) do business with each other (S r ) I ∩ (S p i ) I ≠ Ø SrSr Service Requestor SpiSpi Service Providers Sp1Sp1 SpnSpn... – How do their Service Descriptions intersect ?

8. Slide 8 Overview Introduction Intuition behind modelling service semantics Operationalising discovery using logic Matching service descriptions Conclusion

9. Slide 9 Intuition ↦ DL Service Description ↦ set of DL axioms D ={  1,...,  n } – A service concept S occurring in some  i Domain Knowledge ↦ DL knowledge base KB

10. Slide 10 Intuition ↦ DL Possible World ↦ Model I of KB ∪ D Service Instance ↦ relational structure in I acceptable Service Instances ↦ Extension S I of S Variance due to intended diversity ↦ | S I | ≥ 1 Variance due to incompl. knowl. ↦ several Models I 1, I 2,... Matching ↦ boolean function match( KB, D r, D p ) – way of applying DL inferences (Sr)I1(Sr)I1 (UKCity) I 1 (City) I 1 item from (Package) I 1 (Sr)I2(Sr)I2 (UKCity) I 2 (City) I 2 item from (Package) I 2...

11. Slide 11 Towards Intuitive Modelling Primitives Characterising Property Restrictions Multiplicity – single-valued – multi-valued Variety – fixed value – value range Availability – Mandatory – obligatory Range Coverage – Covering – non-covering

12. Slide 12 Overview Introduction Intuition behind modelling service semantics Operationalising discovery using logic Matching service descriptions Conclusion

13. Slide 13 Treating Variance in Matching Resolving Incomplete Knowledge –  holds in every possible world : Entailment KB ∪ D r ∪ D p ⊨  –  holds in some possible world : Satisfiability KB ∪ D r ∪ D p ∪ {  } sat. Resolving Intended Diversity – Request and Capability overlap : Non-Disjointness  = S r ⊓ S p ⋢ ⊥ – Request more specific than Capability : Subsumption  = S r ⊑ S p – Capability more specific than Request : Subsumption  = S p ⊑ S r ⊨ ⊨   sat. ⊑ ⊓

14. Slide 14 DL Inference for Matching Satisfiability of Concept Conjunction (S r ⊓ S p ) is satisfiable w.r.t. KB ∪ D r ∪ D p (Sr)I1(Sr)I1 (Sp)I1(Sp)I1... (Sr)I2(Sr)I2 (Sp)I2(Sp)I2 ⊨ ⊨   sat. ⊑ ⊓ X (S r ) I ∩ (S p ) I ≠ Ø in some possible world Intuitiuon: – incomplete knowledge issues can be resolved such that request and capability overlap

15. Slide 15 Satisfiability of Concept Conjunction Example: ⊨ ⊨   sat. ⊑ ⊓ X match( KB, D r, D pA ) = true match( KB, D r, D pB ) = true – UKCity ⊓ USCity ⊑ ⊥ is not specified in KB (Sr)I(Sr)I (UKCity) I (City) I Plymouth Dublin from (S pA ) I (USCity) I (S pB ) I (S r ⊓ S p ) is satisfiable w.r.t. KB ∪ D r ∪ D p

16. Slide 16 DL Inference for Matching Entailment of concept subsumption KB ∪ D r ∪ D p ⊨ S r ⊑ S p (S r ) I  (S p ) I in every possible world Intuition: – the request is more specific than the capability regardless of how incomplete knowledge issues are resolved ⊨ ⊨   sat. ⊑ ⊓ X (Sr)I1(Sr)I1 (Sp)I1(Sp)I1... (Sr)I2(Sr)I2 (Sp)I2(Sp)I2

17. Slide 17 Entailment of Concept Subsumption Example: match( KB, D r, D pA ) = false – Dublin outside the UK ⊨ ⊨   sat. ⊑ ⊓ X (Sr)I(Sr)I (UKCity) I (City) I Plymouth Dublin from (S pA ) I KB ∪ D r ∪ D p ⊨ S r ⊑ S p

18. Slide 18 DL Inference for Matching Entailment of Concept Non-Disjointness KB ∪ D r ∪ D p ⊨ S r ⊓ S p ⋢ ⊥ (S r ) I ∩ (S p ) I ≠ Ø in every possible world Intuition: – the request and the capability overlap regardless of how incomplete knowledge issues are resolved (Sr)I1(Sr)I1 (Sp)I1(Sp)I1... (Sr)I2(Sr)I2 (Sp)I2(Sp)I2 ⊨ ⊨   sat. ⊑ ⊓ X

19. Slide 19 Entailment of Concept Non-Disjointness Example: match( KB, D r, D pA ) = true match( KB, D r, D pA ) = false – Plymouth outside the US in at least one possible world ⊨ ⊨   sat. ⊑ ⊓ X (Sr)I(Sr)I (UKCity) I (City) I Plymouth Dublin from (S pA ) I (USCity) I (S pB ) I KB ∪ D r ∪ D p ⊨ S r ⊓ S p ⋢ ⊥

20. Slide 20 Practicability of Inferences Satisfiability of Concept Conjunction – very weak : vulnerable to false positive matches – relies on additional disjointness constraints in domain ontologies Entailment of Concept Subsumption – Very strong : misses intuitively correct matches Entailment of Concept Non-Disjointness – Tries to overcome deficiencies of the other two inferences – relies on range-covering property restrictions (problematic to express in DL)

21. Slide 21 Ranking Service Descriptions Ranking based on Partial Subsumption (SpA ⊓ Sr)I(SpA ⊓ Sr)I DL Inference KB ∪ D r ∪ D p A ∪ D p B ⊨ (S p A ⊓ S r ) ⊑ (S p B ⊓ S r ) ⇒ DpA ≼ DpBDpA ≼ DpB (SpB ⊓ Sr)I(SpB ⊓ Sr)I (SpA)I(SpA)I (SpB)I(SpB)I

22. Slide 22 Conclusion Provided an intuitive semantics for formal Service Descriptions based on Service Instances Emphasized the meaning of variance in Service Descriptions Mapped intuitive notions to formal elements in DL Investigated different DL inferences for matching Service Descriptions Showed how variance can be treated during matching Proposed a ranking mechanism based on partial subsumption of Service Descriptions